Abstract

The power of four cardiorespiratory symptoms to predict subsequent mortality has been analysed in data derived from a random sample of the population of Great Britain recruited in 1965 and followed for 12.4 years. The associations of respiratory symptoms with all causes of death (except stroke) and of cardiovascular symptoms with death from coronary heart disease were strong. The trends of these two associations over the 12.4 years of the follow-up differed substantially: the relative death rates associated with respiratory symptoms remained throughout at about the same level, while those associated with cardiovascular symptoms declined after four years. The excess premature deaths associated with presence of one or more symptoms at entry represented about a quarter of the observed deaths of men and one tenth of those of women.